# Exploring the Validity of the Velocity Matters Linear Position Transducer in the Back Squat and Bench Press

## TL;DR

This study tested a new device for measuring barbell velocity during squats and bench presses and found it has limited accuracy for bench presses.

## Contribution

The study provides empirical validation of the Velocity Matters linear position transducer against a reference standard in two common weightlifting exercises.

## Key findings

- Velocity Matters showed good to excellent correlations with the reference standard across velocity ranges.
- The device systematically underestimated velocity, especially in bench press exercises.
- It may be cautiously used for monitoring mean velocity in squats at lower velocities but not for bench presses.

## Abstract

The purpose of the present study was to validate a new linear encoder by comparing the mean velocity (MV) and peak velocity (PV) of two linear position transducers during free-weight back squat (SQ) and bench press (BP) exercises. Barbell velocity was simultaneously recorded using GymAware (version 5.1.0; reference standard) and Velocity Matters. Fifteen male participants completed two testing sessions, each involving six repetitions (two sets of three) across five velocity ranges: >1.00 to 0.51 m·s−1 (velocity range 1: >1.00 m·s−1; velocity range 2: 0.87–0.99 m·s−1; velocity range 3: 0.75–0.86 m·s−1; velocity range 4: 0.63–0.74 m·s−1; velocity range 5: 0.51–0.62 m·s−1) in SQ and >1.02 to 0.40 m·s−1 (velocity range 1: >1.02 m·s−1; velocity range 2: 0.86–1.01 m·s−1; velocity range 3: 0.70–0.85 m·s−1; velocity range 4: 0.56–0.69 m·s−1; velocity range 5: 0.40–0.55 m·s−1) in BP. In total, 180 repetitions per velocity range were analyzed for each exercise. Validity was assessed using Pearson’s correlation (r), mean absolute error (MAE), Bland–Altman plots, the intraclass correlation coefficient (ICC), and the concordance correlation coefficient (CCC). Pearson’s r indicated good (0.5–0.7) to excellent (>0.9) correlations across all ranges and exercises. However, acceptable MAE values were found only for MV in SQ (except at >1.00 m·s−1) and for both MV and PV in BP at velocities <0.70 m·s−1. Despite an acceptable MAE in some cases, Bland–Altman analyses revealed systematic underestimation by Velocity Matters, with wide limits of agreement of up to −0.08 m·s−1 in SQ and −0.09 m·s−1 in BP, even where MAE was acceptable. ICC values were generally >0.70 but showed wide confidence intervals, indicating high uncertainty. CCC values were consistently poor (<0.90) across all velocity ranges and both exercises, except for PV in the lowest velocity range during BP. In conclusion, Velocity Matters may be cautiously used to monitor MV during SQ at velocities <0.87 m·s−1, but it does not provide sufficient accuracy for use in BP across any load.